Geochemical characteristics of sediment- and mafic rock-hosted Cu deposits in the Kåfjord area, Alta-Kvænangen Tectonic Window, Northern Norway

(1)

Faculty of Science and Technology Department of Geosciences

Geochemical characteristics of sediment- and mafic rock-hosted Cu deposits in the Kåfjord area, Alta-Kvænangen Tectonic Window, Northern Norway

Sondre Stenvold Simonsen

Master’s thesis in Geology, GEO-3900, June 2021

(2)

(3)

Abstract

The Precambrian Alta-Kvænangen Tectonic Window (AKTW) located in the northern part of Norway, hosts different types of copper mineralization. Some of Cu occurrences have been previously mined, but their geochemical and stable isotope characteristics have not been a subject of detailed investigations, and therefore the relevant ore-forming processes are still poorly understood. This study, brings new mineralogical, geochemical and stable isotope data collected from the sediment-hosted and the mafic rock-hosted Cu mineralization in the

Kåfjord area of AKTW.

The Kåfjord area hosts Cu occurrences in sedimentary rocks of the Storviknes formation and in mafic rocks of the Kvenvik formation. Both formations are regionally folded, with

Storviknes formation lying stratigraphically above Kvenvik formation. The Cu mineralization is mostly related to epigenetic quartz-carbonate veins. Considering their mineral assemblages and spatial relationship with the host rocks, the veins were subdivided into 5 different types.

In the three types of sediment-hosted Cu mineralization, Cu occurs in digenite, bornite and/or chalcopyrite. In contrast, in mafic rock-hosted Cu mineralization, chalcopyrite is the only Cu- bearing mineral. In addition to the epigenetic mineralization, a syngenetic mafic rock-hosted VMS mineralization has also been identified.

The mafic rock-hosted quartz-carbonate veins show positive δ¹³C values, indicating

influences from ¹³C-rich carbonate layers formed during the Lomagundi-Jatuli Event, while the δ¹³C and δ¹⁸O values of the sediment-hosted rocks shows a marine origin. The δ³⁴S values of sulfides from mafic rock-hosted veins indicate an influence of evaporites.

The fluid inclusion study revealed that the ore-forming fluid was highly saline with a low to moderate temperature. The high salinity controlled the capability of the fluid to transport Cu in the form of chloride complexes. The same type of highly saline fluid inclusions has been found in the sediment- and mafic rock-hosted Cu mineralization, suggesting that both types of mineralization are a product of the same ore-forming event. The study area displays

geochemical characteristics typical for classical examples of sediment-hosted Cu deposits, with the Storviknes formation identified as the high-grade zone of such deposit, and the mafic rock-hosted Cu mineralization within Kvenvik formation as the low-grade zone. The mafic rocks within the Kvenvik formation are most likely the source of copper.

(4)

(5)

Acknowledgements

This project was funded by the MinExTarget project, EiT RawMaterials.

I will first of all give a big thanks to my supervisor Sabina Strmić Palinkašand co-supervisor Harald Hansen, for all help with this thesis that would not have become any thesis without your help.

Thanks to Yulia Mun for all help and to Fredrik Sahlström for long days in the electron microscope lab. Also, a big thanks to the staff at the geological lab of UiT, to Trine, Ingvild, Karina and Matteus for analysis and guidance through sample preparations. Thanks to the University of Bergen for whole rock analysis, the University of Lausanne and the Stable Isotope Laboratory at CAGE for isotope analyses, and Hugh at GTK for laser ablation analyses. Thanks to Eirik Stokmo, especially for the help during the fieldwork where your SUV came along well.

Thanks to all the people at the red barrack (and specially Torgrim for many coffee breaks) for an enjoyable year, and a special thanks to my office partner Johan, for a good collaboration through this year with valuable talks and discussions.

(6)

1 Introduction

1.1 Context of the study

This master project is a part of a larger project entitled “MinExTarget” (Enhanced Use of Heavy Mineral Chemistry in Exploration Targeting). The project is founded by EIT (European Institute of Innovation and Technology) RawMaterials (EIT RawMaterials, 2020a). The aim of EIT RawMaterials is to: “Enable sustainable competitiveness of the European minerals, metals, and materials sector along the value chain by driving innovation, education, and entrepreneurship” (EIT RawMaterials, 2020b).

MinExTarget is oriented to the early stages of the mining value chain and covers the field of geology and mineral exploration (Figure 1). Therefore, the main goal of MinExTarget is to develop a new tool that can be used in mineral exploration to easier locate undiscovered mineral resources. The new tool is based on the concept that the targeting and qualifying of primary sources of mineralogical and geochemical anomalies can be done faster and more precisely by looking at the trace element concentration and the stable and radiogenic isotope composition of heavy minerals (MinExTarget, 2020).

Figure 1: The mining value chain.

This master thesis is one of two theses that test the MinExTarget concept on the Cu

mineralization in the Alta-Kvænangen Tectonic Window (AKTW), Northern Norway (Figure 3). This thesis is focused on mineral, geochemical and stable isotope characterization of the Cu mineralization, while the thesis by Hilmo (2021) gathers the mineral, geochemical and stable isotope data from stream sediments in streams that drain the Cu mineralization in the study area.

Geology Exploration Exploitation Mineral

processing Environmental

issues Social

aspects

(10)

1.2 Purpose of the study

The main goal of this master project is to determine the mineralogical, geochemical, and stable isotope characteristics of the ore mineralization, host rocks, and alteration products of the sediment-hosted and mafic rock-hosted Cu mineralization in the Kåfjord area of AKTW (Figure 3). The study combines transmitted and reflected polarized light microscopy, scanning electron microscopy coupled with an energy dispersive system (SEM-EDS), lithogeochemistry, carbon (δ¹³C), oxygen (δ¹⁸O) and sulfur (δ³⁴S) isotope analyses, and a fluid inclusion study. In addition, laser ablation inductively coupled plasma mass

spectrometer (LA-ICP-MS) has been used to identify trace element signatures of selected sulfides and oxides.

1.3 Mining history

The main outcome of the mining history is based on the book of Abrahamsen and Veiseth (1997) and Moberg (1968). Other sources are referred to in the text.

Kåfjord in Alta is known for its mining history of copper that started back in the 1800^th century. It has produced thousands of tons of copper from localities in Kåfjord but also from localities farther away in Kvænangen and Raipas (Figure 2).

Figure 2: (A) An orthophoto showing Kvænangen, Kåfjord and Raipas laying within the AKTW (indicated by the white dashed lines). (B) Location of AKTW marked as the red square within Norway. Both orthophotos are downloaded and modified from Norge i bilder (2021).

(11)

The oldest information about prospecting in the Kåfjord area dates back to 1690s, but the mining of copper did not start until Alten Copper Mines (later “Alten Copper Works”) was established by the English merchants Henry Woodfall and John Rice Crowe in 1826. These merchants opened the mine due to promising results from another Englishman called Mitchell, who had some experience in geology and investigated the area the year before on behalf of the merchants and wrote: “the whole of this Alpine district is impregnated with Copper” (Abrahamsen & Veiseth, 1997, p. 4). In 1833, 100 workers worked in the mines of Kåfjord, and a total of 426 persons were living there. Forty-two years later, in 1875, there was a total of 2 360 persons living in Kåfjord. This mining operation was later driven by other English directors after Crowe and Woodfall ended their effort respectively in 1940 and 1944, until it was temporarily closed in 1878 due to several reasons. Among other of these reasons is described by Moberg (1968) as a general loss of energy among the people in the

community. This was the main cause that ended this mining period, but trouble with infill of water in the mines and trouble of removing material as the deeper the mines got should also be pointed out, together with bad management at that time, as the mine operation was leaded from London.

In 1896, the Swedish Nils Persson bought the operation and started mining again. He also disassembled the old melting house and built a new melting house. While disassembling the old melting house, he found 32.5 tons of copper under a furnace which paid off both the rights to mine and the copper processing plant. Electricity was introduced in 1903, generated by waterpower from Møllneselva. This operation lasted until it was shut down in 1909 due to lower copper grade. Also, most known mines started to run out of material, and only smaller new deposits were discovered. Lower copper prices may also have had an effect, and this is the last period of mining in Kåfjord as of today’s date (Mørk, 1970). A total of at least 32 mines have produced copper in Kåfjord, including the mines opened during the first mining period.

Further west of the Kåfjord mines, there are, according to Zwaan and Gautier (1980), two mines called Anna and Lundstrøm that are mineralized. Any other information about the mining operation of these mines is not known, but it is likely to assume that these mines were mined simultaneously with the Kåfjord mines, or later due to their more remote location (Figure 3; Figure 4).

(12)

The processing plant in Kåfjord did also process rich copper ore material from mines in Kvænangen and Raipas during both periods of mining, also owned by Alten Copper Works.

In the winter of 1851, 30 horses transported 12-15 000 kg of ore material by sleds from Kvænangen to Kåfjord over the mountain, but this was too ambitious and was later transported around the mountain.

To get the ore body into loose rocks, iron drills and explosives were used, and then the ore was detached from the barren rocks by crushing by hand (banking). This was not harmless work, and several people died in accidents related to the explosives. The ore material was then transported to the plant, where it was crushed by a crusher and then shipped by boat to England to melt it. From 1835(-36), the ore material was melted in Kåfjord as a melting house was built, and in 1836, a wet separator was installed to separate ore material from barren material after crushing it. During the second period of mining, the barren material was cast in rectangular blocks and used as a building material. These blocks are for instance covering the entrance of Kåfjord church today.

During the very first period of mining in Kåfjord, from 1827 to 1833 there were mined 3 644 tons of ore material with a value of 113 502 speciedalar, according to Moberg 1968. For the period 1843 to 1878, there were mined 61 947 tons of ore that produced 3 022 tons of copper with an average Cu grade of almost 4.9%. Of this material, “Gamla gruvan” delivered the very most material with 55 690 tons of ore material. For the second period of mining, from 1896-1909, there was produced 1 650 to 1 700 tons of copper. In total, there may have been extracted up to 6 000 tons of copper from the copper mines in Kåfjord. In addition,

Kvænangen produced 6 800 tons of ore material with an average grade of 7.4% copper, and Raipas 12 500 tons with an average grade of 6.74% copper during this second period of mining, giving 1 350 tons of copper.

(Moberg, 1968; Abrahamsen & Veiseth, 1997)

(13)

1.4 Samples and methods

Field sampling

Field-work took place in the Kåfjord area (Figure 2) from 10^th to 20^th of August 2020 with the main goal to collect representative samples of the sediment-hosted and mafic rock-hosted Cu mineralization and their host rocks. At the same time, stream sediments from the streams that drain the Cu mineralization were sampled (Hilmo, 2021). In total, 78 rock samples were collected at 20 different localities. Seven of those localities were the historical mines

“Innerstrømmen”, “Carl Johan”, “Wilson”, “Mitchell”, “Henning”, “Anna” and “Lundstrøm”, which can be seen in Figure 3. To get an accurate location of the different samples that were collected, a Garmin GPSMAP 64st was used. This GPS has a quad-helix antenna and a high sensitive GPS and GLONASS receiver that gives a precision of 5 to 10 meters (Garmin, 2020a, 2020b).

Sample selection

Thirty-two rock samples at 14 localities were selected for further mineral, geochemical and stable isotope analysis (Table 1). These localities include the historical mines and also other localities called; “Melsvik”, “Melsvik tunnel”, “Kråknes”, “Kåfjord bridge”, “Carbonate wall”, “Skoadduvarri” and “Møllnes river”, with the location of these localities and selected samples presented in Figure 3.

(14)

Table 1: All samples used in this thesis with corresponding localities, location and information about which investigation methods that have been used. Abbreviations:

Carbonate isotope; Carb, Sulfur isotopes; Bn: Bornite, Dg: Digenite, Ccp: Chalcopyrite, Py: Pyrite.

Locality Sample Location (UTM 34W) Thin section Fluid inclusion Lithogeochemistry SEM-EDS LA-ICP-MS Carbonate isotope Sulfur isotope Isotope sample

Skoadduvarri 030 577555mE 7762159mN x

001 576831mE 7769295mN x 001(carb)

0012 576834mE 7769291mN x 0012(carb)

Møllnes river 045 577539mE 7762469mN x 045(carb)

0581 577350mE 7760588mN x3 0581A(carb), 0581B(carb), 0581C(carb)

042 577341mE 7760576mN x 042(carb)

043 577341mE 7760576mN x 043(carb)

L2 575996mE 7763254mN L2 x x x2 x L2(bn + dg), L2A(carb), L2B(carb)

L1 575996mE 7763254mN L1 x

032 576051mE 7763339mN 032A, 032B x (032B) x x 032A(carb), 032B(ccp)

034 576006mE 7763267mN x

025 575339mE 7761194mN x3 025A(carb), 025B(carb), 025C(carb)

0221 575339mE 7761194mN 0221 x x x x x2 0221A(bn + minor ccp), 0221B(carb), 0221C(ccp + minor bn)

0222 575339mE 7761194mN 0222 x x2 x 0222A(carb), 0222B(ccp)

036 575586mE 7761410mN

037 575611mE 7761379mN x 037(carb)

038 575615mE 7761370mN x 038(carb

039 575632mE 7761371mN x 039(carb)

040 575427mE 7761092mN 040 x x x

0045 580117mE 7767335mN x x 0045A(carb), 0045B(ccp)

0046 580063mE 7767641mN x

Melsvik tunnel 003 577581mE 7768308mN x

047 578676mE 7762997mN 047 x x x2 047A(py), 047B(py + minor ccp)

051 578676mE 7762997mN 051 x x x2 051A(py + ccp), 051B(py + ccp)

057 577532mE 7760660mN 057 x x2 057(ccp + minor py), 057B(ccp + minor py)

0571 577532mE 7760660mN 0571 x x x2 0571A(carb), 0571B(py + ccp), 0571C(py + ccp)

Wilson 020 577207mE 7759773mN x 020(carb)

013 577357mE 7759838mN 013A, 013B x x2 x2 x2 x 013A(py + minor ccp), 013B(carb), 013C(carb)

014 577357mE 7759838mN 014 x x 014(carb)

Innerstrømmen 060 576958mE 7758582mN 060A, 060B x (060A) x (060A) x3 060A(ccp), 060B(ccp + minor py), 060C(ccp)

0051 578334mE 7761115mN 0051 x x x x2 0051A(py), 0051B(carb), 0051C(py)

0052 578370mE 7761187mN x x 0052(carb)

0054 578334mE 7761115mN 0054 x x x

0055 578334mE 7761115mN

Total 32 18 5 4 12 14 28 20

Mitchell

Carl Johan

Kåfjord bridge Melsvik

Carbonate wall

Lundstrøm

Anna

Kråknes

Henning

(15)

Figure 3: Orthophoto of the study area with the location of studied Cu mineralization and collected samples. The orthophoto is downloaded and modified from Norge i bilder (2021).

(16)

Figure 4: A 1:50 000 bedrock map from (NGU, 2021b) overlain an orthophoto from Norge i bilder (2021) with red circles indicating locations of the studied Cu mineralization. The synclinal and anticlinal trend has been added after Bergh and Torske (1988).

(17)

Figure 5: Aeromagnetic total field map of the main study area (NGU, 2021c) with localities marked by red circles.

The thick black line marks the transition to the sea. Localities (as Wilson, Carl Johan, and Kåfjord bridge) within the mafic rocks of Kvenvik formation show a highly positive anomaly, while localities (as Anna and Lundstrøm) within the sedimentary rocks of the Storviknes formation shows a highly negative anomaly.

Sample preparation

The sample preparation took place in “Emilbua” at the Department of Geosciences of UiT. A MK-101 wet tile saw with a diamond blade was used for cutting the selected samples, into smaller pieces suitable for preparation of thin or thick sections, or just to open rocks to expose the texture.

(18)

A rock crusher was used for crushing four samples into <5 mm fragments for

lithogeochemical analysis. Before the samples were crushed, they were cut into suitable pieces by the wet saw, and unwanted parts as mineralized or weathered parts were cut off.

The crushed material was milled in a Retsch PM 100 agate miller, into a very fine-grained homogenous powder for lithogeochemical analysis.

As the carbonate and sulfur isotope analysis only requires 0.1 grams of pulverized material, a Dremel 3000 drill machine with a diamond drill head was used to extract powder of carbonate and sulfide minerals by drilling. Before drilling each sample, the drill head was cleaned by first using a paper towel to remove most of the substances. Then, the drill head was dipped in 10% hydrochloric acid (HCl) as the remaining carbonates will react with HCl and form water, carbon dioxide, and calcium chloride (CaCO3(s) + 2HCl(aq) → H2O(l) + CO2(g) +

CaCl2(aq)). The drill head was then dried and cleaned with towel paper to ensure no substances were left on the drill head. For each drilled sample, a new paper sheet was laid underneath as a base to catch the powder and to avoid contamination. Each sample of drilled powder was put in small glasses and sent to the Stable isotope laboratory of Institute of Earth Surface Dynamics, University of Lausanne, Switzerland, for analyses of carbon (δ¹³C),

oxygen (δ¹⁸O), and sulfur (δ³⁴S) composition. In addition, two samples of sulfides (δ³⁴S) were prepared for stable sulfur isotope analysis and delivered to the Stable Isotope Laboratory at CAGE – Centre for Arctic Gas Hydrate, Environment and Climate, Department of

Geosciences of UiT.

Polished section preparation

The rock slabs, cut in approximately 1 x 2 x 3 cm dimensions, from selected fifteen rock samples were delivered to the geological laboratory at the Department of Geosciences, UiT, for further preparation of polished thin sections (30 μm thickness) for transmitted and reflected polarized light microscopy, SEM-EDS and LA-ICP-MS analyses. Polished thick sections (300 μm thickness) from selected six rock samples were prepared for fluid inclusion study.

Preparation for fluid inclusion study

Preparation of the double polished wafers was performed in the geological laboratory at the Department of Geosciences, UiT. Selected samples were cut into approximately 1 x 2 x 3 cm slabs. One 2 x 3 cm surface of each slab was grinded by a Logitech LP-50 with silicon carbide blades for 30 minutes and polished in a Phoenix Beta polishing machine using 6, 3

(19)

and 1-micron diamond pastes and suitable polishing cloths. The machine was set to a rotation speed of 75 rpm while the sample was manually held on the fiber mash and turned against the rotation direction of the plate. Samples containing quartz were polished for 25-27 minutes (15 minutes with 6-micron diamond paste, 7-8 minutes with 3-micron diamond paste and 3-4 minutes with 1-micron diamond paste). Samples containing minerals softer than quartz were usually polished 15-17 minutes (10 minutes with 6-micron diamond paste, 3-4 minutes with 3-micron diamond paste and 2-3 minutes with 1-micron diamond paste). The quality of the polishing was regularly checked in a light microscope and an ultrasonic bath was also regularly used to remove particles from the surface of the wafers.

In the next step, the slab was mounted onto a standard microscopy glass with the polished part towards the glass. Crystal bond was used as an adhesive and the process was performed on a hot plate at a temperature of 120°C. After fixation, the slabs were first cut with a Struers discoplan TS wet saw and then grinded with a Struers discoplan TS grinder to the final thickness. The final thickness is usually around 250 μm but it may vary from sample to sample depending on its transparency, size of entrapped fluid inclusions, as well as on the distribution of fluid inclusion assemblages (Goldstein, 2003).

When correct thickness was obtained, the unpolished side of the sample had to be grinded by hand, by using silicon carbide with a grain size of 600K on a piece of glass as a base for about 5 minutes with water as lubricant. After the grinding was done, all the samples were polished again using the same procedure for polishing as mentioned before by the 6-micron, 3-micron and 1-micorn fiber mesh with associated diamond paste.

As a final step of the process, the glass had to be removed from the samples. To do this it was put on the hot plate again and warmed up until the crystal bond melted. The sample was quickly slid off the glass into a bowl with acetone. This bowl of acetone with the thick section inside was then laid in a supersonic bath (floating on the water), and it was runned for 2 minutes as the remaining glue from the crystal bond was dissolved by the acetone. Sample by sample was checked for remaining glue and removed by spraying on acetone if necessary until all glue was removed, and the sample was clean and ready for fluid inclusion study.

(20)

Analytical methods

1.4.4.1 Polarized light microscopy

Characterization of micro-textures as well as identification of mineral phases were acquired from polished thin and thick sections using a Leica DMLP microscope. Transmitted light microscopy (TLM) and reflected light microscopy (RLM) were applied in the identification of transparent and opaque minerals, respectively. Photomicrographs of micro-textures and characteristic phase relationships were acquired using a Leica DMC4500 camera and the Leica Application Suite software.

1.4.4.2 Whole rock geochemistry (Major and trace element analysis by ICP-MS, ICP-AES and XRF)

The lithogeochemical analyses of four samples (003, 0054, 0046 and 013) were done at the Department of Geosciences of the University of Bergen. To plot the samples in diagrams, the program “R” with the package “GCD Kit” has been used (Janoušek et al., 2006).

Further preparation of samples for ICP-MS, ICP-AES, and XRF

The aliquots of 0.5 g pulverized samples were dried and heated up to 1000°C for two hours to determine the loss of ignition (LOI). For ICP-MS and ICP-AES analysis, 100 mg of the samples were weighted in 25 ml PFA Savillex beakers and digested in 3 ml concentrated HF on a heating plate at 135°C during 48 hours. The HF supernatant was evaporated to dryness and the fluoride residues were subsequently hydrolyzed in a weak solution of HNO3 on the heating plate under sub-boiling point conditions and evaporated to dryness. The resulting nitrate salt residue was dissolved in a ca 2 ml 2N HNO3 prior to dilution with 2% HNO3 in 50 ml in volumetric flasks.

Major and trace element analysis by ICP-AES

A Thermo Scientific ICap 7600 Inductively Coupled Plasma Atomic Emission Spectrometer (ICP-AES) is used to measure the concentration of Al, Ca, Cu, Fe, K, Mg, Mn, Na, P, and Ti.

External calibration curves from Spectrapure (certified single element solutions to prepare multi-element standard solutions) are used to do the quantification. Internal standardization of Sc is used, and samples are diluted by 2% w/v HNO3 before analysis.

A USGS CRM BCR2 (Basalt, Columbia river) is used for quality control, with BCR2 following all steps from digestion. Repeatedly analysis of synthetic water CRM SPS-SW-2

(21)

(Spectrapure Standards AS) is done to monitor the performance during analysis and to control the calibration curves.

Trace element analysis by ICP-MS

A Thermo Scientific Element XR High-Resolution Inductively Coupled Plasma Mass

Spectrometer (HR-ICP-MS) is used to measure trace element concentration of Ce, Co, Cr, Cs, Cu, Dy, Er, Eu, Gd, Hf, Ho, La, Li, Lu, Nb, Nd, Ni, Pb, Pr, Rb, Sc, Sm, Sr, Ta, Tb, Th, Ti, Tm, U, V, Y, Yb, Zn, and Zr. External calibration curves from Spectrapure (certified single element solutions to prepare multi-element standard solutions) are used to do the

quantification. Internal standardization of In is used, and samples are diluted by 2% w/v HNO3 before analysis.

A USGS CRM BCR2 (Basalt, Columbia river) is used for quality control, with BCR2 following all steps from digestion. Repeatedly analysis of synthetic water CRM SPS-SW-2 (Spectrapure Standards AS) is done to monitor the performance during analysis and to control the calibration curves.

Major element analysis by XRF

A Bruker S4 PIONEER X-ray fluorescence spectrometer is used to analyze the concentration of Si. External calibration curves from USGS based on several powdered CRMs are used for quantification. Analysis of USGS CRM BCR2 (Basalt, Columbia river) is used for quality control.

1.4.4.3 Mineral chemistry by SEM-EDS

A Zeiss Merlin Compact VP field emission scanning electron microscope (FE-SEM) with a X-Max energy dispersive system (EDS) at the Faculty of Health, UiT, was used to identify the ratio between elements in different minerals in weight percent (wt. %). Prior to the SEM- EDS analyses, polished sections were carbon-coated (10 nm) using a Leica EM ACE600 sputter coater (the first batch of analyzed samples) and a Quorum 150R ES plus coater (the second batch of analyzed samples).

A working distance (WD) of 8,5 mm between the EDS detector and the sample was used with an electron high tension of 15 kV on the first batch (thin section 040, 051, 0054, 0221, L1 and L2) and 20 kV on the second batch of samples (thin section 013A, 013B, 040, 047, 0051, 057, 060A and L1). Corrections of elements were done using Aztec analysis software from Oxford Instruments.

(22)

1.4.4.4 Trace element analyses by LA-(SC)-ICP-MS

Trace element analysis of specific minerals was done in Finland at the Geological Survey of Finland (GTK), using an Analyte 193 ArF (Photon Machines, San Diego, USA) laser-ablation (LA) system in combination with a Nu AttoM (Nu Instruments Ltd., Wrexham, UK) single collector inductively coupled plasma mass spectrometer (SC-ICP-MS).

The laser was run at a pulse energy of 5 mJ at 30% attenuation and a pulse frequency of 5 Hz to produce an energy flux of 2.17 J/cm2 on the sample surface with a 25 μm spot size for oxides and a 40 μm spot size for sulfides. These two spot sizes were chosen to provide the best compromise between adequate resolution, to allow spot analysis of compositional zones determined by high contrast SEM imaging, still keeping limits of detection (LOD) as low as possible. Each analysis was initiated with a 20-second baseline measurement followed by switching on the laser for 40 seconds for signal acquisition. Analyses were made using time- resolved analysis (TRA) with continuous acquisition of data for each set of points (generally following the scheme of primary standard, quality control standard, 15 unknowns). For the analyses of oxide minerals (magnetite, hematite, ilmenite, titanite, rutile) GSE glass was used as the primary external standard, with GSD glass BHVO-2G and BCR-2G as reference materials for quality control. For the analyses of sulfide minerals (pyrite, chalcopyrite, sphalerite, bornite, digenite, tennantite, molybdenite) UQAC FeS-1 was used as the primary external standard, with USGS MASS1 as reference material for quality control. The isotope

57Fe has been used as an internal standard. The measurements were performed on 34 isotopes for oxides and 40 isotopes for sulfides covering 34 elements for oxides and 38 elements for sulfides at low resolution (∆M/M = 300) using the fast scanning mode. Data reduction was handled using the software GLITTER TM (Achterberg et al., 2001) which allows the baseline subtraction, the integration of the signal over a selected time resolve area, and the

quantification using known concentrations of the external and internal standards (GTK, 2021).

1.4.4.5 Fluid inclusion study

Petrographic and microthermometric measurements of fluid inclusions were performed at the Department of Geosciences of UiT. Six samples were selected from quartz-carbonate veins hosted by sedimentary rocks (samples 030, 034, 040 & 0221), and by gabbro (sample 060 &

0052).

(23)

Microthermometric measurements were carried out on Linkam THMS 600 stages mounted on an Olympus BX 2 microscope using 10× and 50× Olympus long-working distance objectives.

Two synthetic fluid inclusion standards (SYN FLINC; pure H²O and mixed H²O-CO²) were used to calibrate the equipment. The precision of the system was ±2.0°C for homogenization temperatures, and ±0.2°C in the temperature range between –60° and +10°C. The

measurements were made on carefully defined fluid inclusion assemblages (FIAs),

representing groups of inclusions that were entrapped simultaneously from the same fluid.

The fluid inclusion assemblages were identified based on petrography prior to heating and freezing. If all fluid inclusions within the assemblage showed similar homogenization

temperature, the inclusions were assumed to have trapped the same fluid and to have not been modified by leakage or necking; these fluid inclusions would thus record the original trapping conditions (Goldstein & Reynolds, 1994; Goldstein, 2001; Bodnar, 2003a).

During microthermometric measurements, the following phase transitions were recorded: the first-melting temperature (eutectic Te); last melting temperature of hydrate (Thyd); last melting temperature of ice (Tm ice); last melting temperature of halite (Ts); and the total

homogenization temperature (TH). Calculations of compositions, densities, and isochores were conducted applying the numerical model by Steele-MacInnis et al. (2011) for the H2O- NaCl-CaCl2 system and Steele-MacInnis et al. (2012) for the H2O-NaCl system.

1.4.4.6 Stable isotope analyses

Carbon and oxygen isotope analyses of carbonates were performed at the Stable isotope laboratory of Institute of Earth Surface Dynamics, University of Lausanne, Switzerland.

Measurements were carried out with an automated Thermo/Finnigan online preparation device Gas Bench II connected to a isotope ratio mass spectrometer (IRMS) using a

continuous flow mode (Révész & Landwehr, 2002). Borosilicate sample bottles were washed in diluted acid, then twice in deionized water and overnight dried at 70°C. The powder samples (~250μg) were added to the vials in air, and air was removed from the sample vials by automatic autosampler-assisted flushing with He, using He flow of 100 ml/min for 5 minutes. The phosphoric acid, which is maintained at the reaction temperature (70°C for calcite and 90°C for Fe-carbonates and dolomite) was added dropwise under computer control to each individual reaction vessel. The reaction time was 60 minutes. Both the amount of the acid and the reaction time were controlled by the software.

(24)

Data was extracted to an EXCEL file by using the ISODAT NT EXCEL export utility and further calculation steps were carried out using a predefined EXCEL Worksheet. A linearity correction was applied based on the relationships between the intensity of the first sample peak (m/z 44) and ¹⁸O value of the standards. The stable carbon and oxygen isotope ratios are reported in the delta () notation as per mil (‰) deviation relative to the Vienna Standard Mean Ocean Water (V-SMOW) for oxygen and Vienna Pee Dee Belemnite (V-PDB) for carbon. The analytical reproducibility was better than ±0.05‰ for ¹³C and ±0.1‰ for ¹⁸O.

Sulfur isotope analyses were carried out at performed at the Stable isotope laboratory of Institute of Earth Surface Dynamics, University of Lausanne, Switzerland. Measurements were performed by on-line EA-IRMS system consists of a Carlo Erba 1108 elemental

analyzer (EA) coupled with a continuous helium flow interface to the Thermoquest/Finnigan Mat Delta S IRMS. The EA oxidizes all sample compounds under a stream of helium and oxygen by flash combustion in a single oxidation-reduction quartz tube filled with oxidizing (tungsten trioxide) and reducing (elemental copper) agents at 1030°C. Water was removed using anhydrous magnesium perchlorate, and the gases enter a chromatographic column (Poropak QS) for separation of SO2 which is isotopically analyzed by IRMS (Giesemann et al., 1994). The sulfur isotope values are reported in the typical −notation relative to V-CDT standard. The reproducibility, assessed by replicate analyses of the laboratory standard (natural pyrite, +6.1 ‰, synthetic mercury sulfide, +15.5 ‰, barium sulfate, +12.5 ‰ ³⁴S) was better than 0.2 ‰.

Sulfur isotope composition of two additional sulfide minerals was performed at Stable Isotope Laboratory of CAGE, Department of Geosciences, UiT. Approximately 0.1 g of pulverized sample were loaded in Sn-capsules and were combusted in a Thermo Scientific Flash HT Plus (EA) at 1020°C after being weighed. Analysis of the combusted material was done by an IRMS (Thermo Scientific MAT253) instrument with normalization of δ³⁴S to VCDT (Vienna-Canyon Diablo Troilite) by the three international standards S1, S2, and NBS127.

Uncertainty of the instrument by perfectly homogenous mixed material is a standard deviation of ≤0.20 ‰.

(25)

2 Geological setting

2.1 The Fennoscandian Shield

The Fennoscandian Shield (Figure 6), also called the Baltic Shield, located in Norway, Sweden, Finland, and Russia is the northwestern part of the East European Craton (Gorbatschev & Bogdanova, 1993).

Gaal and Gorbatschev (1987) divided the Fennoscandian Shield into three main domains based on the age of formation: 1) the Archean domain; 2) the Svecofennian domain, and 3) the Southwest Scandinavian domain (Figure 6). The Archean domain is built up of rocks formed during the Saamian (3.1-2.9 Ga) and the Lopian orogeny (2.9-2.0 Ga) while the Svecofennian domain was formed during the Svecofennian orogeny (2.0-1.75 Ga). The Southwest Scandinavian domain was formed during the Gothian orogeny (1.75-1.50 Ga).

Later orogeny’s as the Sveconorwegian orogeny (1.25-0.90 Ga), and the Caledonian orogeny (600-400 Ma) have affected the western part of the Fennoscandian Shield with metamorphism and overlaying strata as the Caledonian Nappes (Gaal & Gorbatschev, 1987).

Figure 6: Overview of the Fennoscandian Shield, including the approximate location of AKTW (Alta-Kvænangen Tectonic Window), modified after Lahtinen et al. (2005), based originally on Gaal and Gorbatschev (1987).

(26)

In Norway, the Fennoscandian Shield represents the basement of the overlaying Caledonian Nappes and it is exposed in several “tectonic windows” (Figure 7). These tectonic windows are a result of extension and erosion after the thrusting of the Caledonian Nappes, as nappes have glided back towards the direction of thrusting, and some places revealed the basement below with the help of erosion amplified by land uplift (Fossen et al., 2013, p. 210).

Figure 7: Tectonic windows within the Caledonian Nappes, AKTW is emphasized. Modified from Fossen et al.

(2013)

The oldest rocks of the Fennoscandian Shield are gneisses found in northern Finland. These gneisses, also known as “Siuruagneiss”, have been dated to 3.5 Ga. Zircons isolated from Siuruanagneiss have been dated to 3.73 Ga, suggesting the existence of even older rock in the Fennoscandian Shield (Nordgulen & Andresen, 2013, p. 71).

The Archean domain

The Archean domain is the northeastern-most part of the Fennoscandian Shield and has been divided into three smaller crustal provinces: the Karelian Province, the Kola Peninsula

(27)

Province, and the Belomorian Province (Figure 6; Figure 8). The majority of the Karelian Province is granite-greenstone belts, while the Kola and Belomorian Province are high-grade gneisses (Gaal & Gorbatschev, 1987).

The Karelian Province underwent several events of rifting between 2.5-1.95 Ga. It started as intracratonic rifts and developed into the opening of basins as “Kola ocean”, with the

development of island arcs. This rifting resulted in volcanic activity with simultaneous erosion from the continent and sedimentation in the evolving basins, which made sequences of alternating volcanic- and sedimentary rocks (Nordgulen & Andresen, 2013, p. 71). The basement of the greenstone belts in the Karelian Province are Saamian tonalitic-

trondhjemitic-granodioritic rocks (Lahtinen et al., 2005).

(28)

Figure 8: Northeastern-most part (Archean domain) of the Fennoscandian Shield with different types and ages of Palaeoproterozoic and Archean rocks. Alta-Kvænangen Tectonic Window (AKTW) is emphasized in the black rectangle. Modified after Melezhik et al. (2013), based originally on Koistinen et al. (2001).

(29)

The Svecofennian domain

The Svecofennian domain (Figure 6) is a terrain formed in the period between 1.92 to 1.79 Ga by the Svecofennian orogeny (Lahtinen et al., 2005). The orogeny occurred due to a period with compression in the western part of the Archean continent (Fennoscandian Shield), where evolved rift basins with volcanic, sedimentary, and intrusive rocks got metamorphosed and folded (Nordgulen & Andresen, 2013, p. 33, 71). Four main stages of orogenic evolution (Nironen (1997) and Lahtinen et al. (2009) within Bogdanova et al. (2015)) are; 1) Accretion of microcontinents (1.92-1.87 Ga); 2) Extension of the continent (1.86-1.84 Ga); 3)

Continent-continent collision (1.84-1.79 Ga), and 4) Orogenic collapse (1.79-1.77 Ga).

The Transscandinavian igneous belt

At the end of the Svecofennian orogeny, huge amounts of granitic magmas started to intrude the western part of the Archean continent in the period from about 1.85 to 1.65 Ga. The remaining granites after this event are called “The Transscandinavian igneous belt” (TIB) (seen in Figure 6), and can be followed from Lofoten in the Northern part of Norway to Skåne in Southern Sweden (Nordgulen & Andresen, 2013, p. 71).

The Southwest Scandinavian domain

The Gothian- and Sveconorwegian orogeny are two periods of orogeneses that occurred respectively 1.75-1.55 Ga (Åhäll & Larson, 2001) and ⁓1.20-0.90 (Gorbatschev &

Bogdanova, 1993) Ga ago in the southwestern part of the Archean continent, where rocks from several different tectonic environments got folded and metamorphosed. These rocks are now found in southern Norway and Sweden, making up the Southwest Scandinavian domain seen in Figure 6 (Nordgulen & Andresen, 2013, p. 71).

Lahtinen et al. (2005) redefined the term “Gothian orogeny” from Gaal and Gorbatschev (1987) to the term “Gothian evolution”, as it includes several events of collisions during a period of about 200Ma.

Alta-Kvænangen Tectonic Window

Alta-Kvænangen Tectonic window (ATKW) is in the northernmost part of the Fennoscandian Shield in northern Norway, where it lies as a tectonic window within the Caledonian Kalak Nappe Complex. The tectonic window has been revealed due to erosion of the overlaying Caledonian Nappes by land uplift, and is deformed in a low degree and metamorphosed at lower facies (Nordgulen & Andresen, 2013, p. 81).

(30)

AKTW consists of Paleoproterozoic sedimentary, magmatic, volcanic, and volcanoclastic rocks formed during rifting that later have been deformed during the Svecofennian orogeny (Figure 12; Nordgulen & Andresen, 2013, p. 71). AKTW hosts the Raipas Supergroup, which Zwaan and Gautier (1980) have divided into four formations. Kvenvik formation is the lowermost known formation which has been divided into Lower- and Upper Kvenvik formation by Vik (1985). Storviknes formation overlays Kvenvik formation, followed by Skoadduvarri formation. Luovosvarri formation completes the supergroup. The base of the Raipas Supergroup is not known, while the Bossekop Group overlays the Raipas Supergroup and appears as a quartzite in the field area (Zwaan & Gautier, 1980).

AKTW has been considered as a northwestern continuation of the Karelian province, whereas aeromagnetic geophysical maps (Figure 9) indicate a continuation from the Kautokeino greenstone belt (KGB) underneath the Caledonides to the AKTW (Melezhik et al., 2015), and also whereas KBG and AKTW show numerous lithological similarities. Both KBG and AKTW, consist partly of metamorphosed basalts and tuff/tuffites deposited in a marine basin, and the Skoadduvarri formation in the Raipas Supergroup in AKTW, is thought to be

equivalent to the Carravarri formation in the KGB (Nordgulen & Andresen, 2013, p. 79, 81).

(31)

Figure 9: Aeromagnetic anomaly map, showing magnetic features from sedimentary-volcanic formations in KGB traced to AKTW underneath the Caledonian Nappe Complex (emphasized by the white dashed lines). The white rectangle indicates the location of the aeromagnetic map in Figure 5. Modified from Melezhik et al. (2015) with

aeromagnetic data obtained from the MINN project database at NGU.

(32)

Figure 10: Stratigraphic column of the Kvenvik, Storviknes and Skoadduvarri formations within AKTW modified from Melezhik et al. (2015), based originally on Vik (1985).

(33)

Raipas Supergroup

2.1.6.1 Kvenvik (greenstone) formation

The Kvenvik formation is the lowermost known sequence of the Raipas Supergroup in AKTW (Figure 10). It consists of an alternating succession of gabbro, tholeiitic MORB-type basaltic lavas, pillow lava, carbonates, tuff, and tuffite, that have a thickness of at least 1500m and are metamorphosed at greenschist facies (Bergh & Torske, 1988). The carbonate has been identified as dolostones according to Melezhik et al. (2015), which divided them into three dolostone sequences, Lower dolostone, Upper dolostone and Uppermost dolostone (Figure 10). Ultramafic bodies can also be found locally in the Kvenvik formation. They are

tectonically re-emplaced rocks and metamorphosed up to the greenschist facies (Gautier et al., 1979). Fifteen repetitive sequences of volcanoclastic rocks and lavas have been identified in the Kvenvik formation (Bergh & Torske, 1988). These sequences have been formed due to multiple eruptions of lava in a terrestrial to shallow-water environment, possibly during an early phase of an intracratonic rift in the Archean basement at the Fennoscandian Shields margin (Bergh & Torske, 1988).

Dating of the Kvenvik formation has been done by two studies. Gautier et al. (1979) have dated fine- and coarse-grained diabases and pillow lava to have a minimum depositional age of 1400 – 1500 Ma by ⁴⁰K/⁴⁰Ardating method. Gautier et al. (1979) also suggest a formation age of 1800 – 2000 Ma for these rocks based on some of their data. Melezhik et al. (2015) have dated the gabbro hosting the Kåfjord deposits to be 2146±5 Ma by the U-Pb dating method of zircons (Figure 10).

2.1.6.2 Storviknes (dolomite) formation

The Storviknes formation is laying stratigraphically above the Kvenvik formation as an unconformity with an 80 Ma non-depositional break and consists of alternating layers of siltstone and dolomites (Melezhik et al., 2015). Laminated structures of stromatolites are found in the dolomite, formed on the oceanic floor when carbonate particles attached to adhesive matters produced from algae and cyanobacteria (Nordgulen & Andresen, 2013, p.

81).

2.1.6.3 Skoadduvarri (sandstone) formation

The Skoadduvarri formation is the upper formation of the Raipas Supergroup (Figure 10), which has a gradual transition from the underlaying Storviknes formation. Most of the

Skoadduvarri formation is sandstone, but siltstones, shales, and conglomeratic sandstones also

(34)

occur and have a total thickness of at least 1700m (Bergh & Torske, 1986). It is thought to have been formed as a delta in a subsiding marine basin due to evidence of tidal plain, turbidite deposits, upward coarsening, crossbedding, wave ripples, and others (Bergh &

Torske, 1986).

2.1.6.4 Luovosvarri formation

The Luovosvarri formation is the uppermost formation of the Raipas Supergroup. This

sequence is approximately 100m thick, consisting of alternating layers of stromatolite bearing dolomite, sandstone, and layers of shale. It is thought to be deposited in a shallow marine setting as there for instance are found wave ripple marks (Zwaan & Gautier, 1980).

Bossekop Group

Bossekop Group overlays the Raipas Supergroup unconformably with a basal layer of sandstone, siltstone, and shale, followed by quartzite (Zwaan & Gautier, 1980).

Ore deposits in the Fennoscandian Shield

The Fennoscandian Shield hosts various types of mineral deposits, including orogenic gold deposits, Ni-Cu±PGE deposits, VMS deposits of Zn-Cu-Pb±Ag±Au, iron oxide-copper-gold (IOCG) deposits, and Fe-Ti deposits (Weihed et al., 2005). The formation of the major orogenic gold deposits has been associated with three time periods, from 2.72-2.67 Ga, 1.90- 1.86 Ga, and 1.85-1.79 Ga, all related to shortening of the crust. Ni-Cu ± PGE deposits have been deposited in seven different geological environments during a period from 2.74 to 1.88 Ga. These are in Archaean (2.74 Ga) and Palaeoproterozioc (2.20-2.05 Ga) greenstone belts, mafic layered intrusions (2.49-2.45 Ga), ultramafic volcanism (1.97 Ga), Palaeoproterozioc ophiolite complexes (1.97 Ga), mafic and ultramafic intrusions (1.88 Ga) and minor deposits related to dykes of diabase formed in post-orogenic regimes. Between 1.97-1.88 Ga, all the major VMS deposits in the Fennoscandian Shield were formed in an extensional regime.

IOCG deposits were formed related to magmatism at 1.88 Ga ago and during the period of 1.80-1.77 Ga. Larger Fe-Ti deposits were formed at 930-920 Ma hosted by anorthositic magma (Weihed et al., 2005).

(35)

3 Theoretical background

3.1 Sediment-hosted copper deposits

Sediment-hosted copper deposits or sediment-hosted stratiform copper deposits, represent copper and copper-iron sulfide mineralization hosted by sedimentary rocks such as

carbonates, siltstones, shales, or sandstones (Figure 11; Kirkham, 1989). The mineralization is often associated with the original bedding of the host rock (Boynton, 1984), but

mineralization can also occur oblique to the original bedding, as sediment-hosted copper deposits can form in different settings and are thought to be of epigenetic origin (Brown, 1997). This means that mineralization is happening after deposition of the host rock, with copper minerals sometimes replaces earlier syngenetic mineralization of iron sulfides (Brown, 1997). The thickness of these mineralized layers is generally less than 30 meters and often less than 3 meters (Hitzman et al., 2010). The deposition most commonly occurr in shallow- marine basins often related to passive continental margins, failed arms of triple junction, or intracontinental rifts with a high rate of evaporation (Cox et al., 2003).

Figure 11: Cross-section of a typical setting of an intracratonic sediment-hosted copper deposit within a closed basin, from Hitzman et al. (2010).

Figure 11 shows a typical sediment-hosted copper deposit in an intracontinental setting with an underlaying basement that is filled with red beds and bimodal (felsic and mafic) volcanic rocks formed during rifting of the continental crust. Marine siltstones, shales, and sandstones that locally can be organic-rich overlays. Marine carbonates with a thick sequence of

evaporites are deposited above these siliciclastic sediments, which are succeeded by

(36)

continental to shallow marine siliciclastic sediments. This sedimentary sequence can have a thickness from a few kilometers up to >10 km, which makes highly saline brines from the evaporitic sequence move down into the oxidized red beds. Convection of the brines is initiated from burial pressure and sometimes igneous activity, whereas this brine-fluid picks up metals from both the red beds (which also oxidizes the fluid) and the underlaying

basement. The metal-rich oxidized brine circulates up and hits organic material, which acts as a reductant in the siliciclastic sediments, and copper sulfides are precipitated. Precipitation of copper can also occur at higher levels by being transported in faults, but in most cases, the evaporitic layer has a function as an uppermost seal of the hydrologic system (Hitzman et al., 2010).

Subtypes

United States Geological Survey (USGS) collected information about 785 sediment-hosted copper deposits around the world and divided this type of deposit into three subtypes (Cox et al., 2003): 1) reduced-facies Cu deposits; 2) redbed Cu deposits; and 3) Revett-type Cu deposits. These three subtypes vary in both the type of reductant rock (often host rock to mineralization), the reducing ability, and the tonnage and the ore grade.

1) Reduced-facies Cu deposits are characterized by an organic matter-rich (high capacity of reducing fluids) fine-grained sediment (e.g. siltstone, shale, mudstone) as a

reductant rock occurring in 41 percent of this type of deposit investigated, often with stromatolites and other tidal evidence from lacustrine or marine origin.

2) Redbed Cu deposits are characterized by having an organic matter-poor (low capacity of reducing fluids) coarse-grained sediment (e.g. sandstone, quartzite or conglomerate) with minor patches of organic debris as reducing medium within the coarser-grained sediment, which is the case for 85 percents of this type of deposit.

3) The Revett-type Cu deposits have a reductant rock that varies and has a low capacity of reducing fluids. The host rock is coarse-grained sedimentary rocks in 77 percent of this type of deposit, whereas sandstone is the most abundant sediment type. In

Phanerozoic deposits, the reducing medium is often fluids of sulfide-rich sour gas or hydrocarbons (Cox et al., 2003).

For the reduced-facies Cu deposits, tectonism is an important factor that drives fluid flow and controls ore deposition (Cox et al., 2003). Some deposits have basalt flows as a source rock of copper as in the sediment-hosted Kennecott-type of deposits in the Wrangler Mountains of

(37)

Alaska. Here, the underlying greenstones are supposed to host the high-grade copper (ore grade of 13 wt.% Cu) occurring as massive CuS-minerals in the border between the

greenstone composed mainly of basalt flows, and the overlying limestone (MacKevett et al., 1997).

Formation criteria’s

Four conditions to form a sediment-hosted copper deposit are required according to Cox et al.

(2003), whereas all of them must be fulfilled to form this type of deposit.

1. The first condition is the presence of an oxidized source rock that is used as a source to leach out copper. This source rock must be a mafic rock, or a rock that contains ferromagnesian minerals such as subaerial volcanic rocks, shales, conglomerates, or red sandstones, and it must also be hematite stable. Marine volcanic rocks are therefore most likely to be unsuitable as a source rock as it has not degassed the volatiles and contain therefore reduced sulfur, which excludes the formation of a hematite-stable environment. The marine volcanic rocks are also not exposed to oxygen in the same degree as subaerial volcanic rocks.

2. The second condition requires a media to mobilize the copper, and this is often brines or evaporites that pick up and transports copper by chloride complexes as they are enriched in sodium and chlorite. Leaching of copper from an oxidized source rock with chlorite complexes can be seen in Equation 1 below.

𝐶𝑢₂𝑂(𝑆) + 6𝐶𝑙⁻(𝑎𝑞) + 2𝐻⁺(𝑎𝑞) → 2𝐶𝑢𝐶𝑙₃²⁻(𝑎𝑞) + 𝐻₂𝑂(𝑙)

Equation 1: Copper transported by chlorite complexes.

3. The third condition requires a reduced fluid or rock to precipitate copper. This fluid rich in reduced sulfur can be formed by sulfate-reducing bacteria under anaerobic conditions that are commonly produced in carbonate rocks or organic-rich shales. The reduced host rocks contain often disseminated mineralization of pyrite as fluids in equilibrium with pyrite also can form reduced fluids. Chalcocite is produced when reduced sulfides react with chlorite complexes, as seen in Equation 2 below.

2𝐶𝑢𝐶𝑙₃²⁻(𝑎𝑞) + 𝑆²⁻(𝑎𝑞) → 𝐶𝑢₂𝑆(𝑠) + 6𝐶𝑙⁻(𝑎𝑞)

Equation 2: Formation of chalcocite from chlorite complexes

4. The fourth condition is that there must be a condition for fluids to mix in the host rock, which primarily is a sedimentary rock, but sometimes this mixing happens in

(38)

volcanic rocks. The host rock needs to have a good ability to transport fluids (high permeability) to allow fluid mixing, which often is related to folding, faults, and graben structures, making way for fluid to circulate and interact. Compaction is also an important factor for fluid mixing. Less permeable shales within the host rock promote conditions for bedding-parallel mixing of fluids and related bedding-parallel mineralization (Cox et al., 2003).

Circulation of fluids in sediment-hosted copper deposits due to a salinity contrast is supposed by Koziy et al. (2009) to be an important mechanism of fluid mixing and transport of deeper- seated copper from igneous basement rocks. This was tested in the Zambian Copperbelt as copper from underlying red beds has been thought to not be a sufficient source of copper by Hitzman (2000), due to too small amounts of and too low copper grade in the available red beds. Koziy et al. (2009) found that even in low permeable rocks, fluid circulation sufficient to leach out elements in igneous basement rocks would occur due to this salinity contrast.

When the effect of salinity was removed, the fluid circulation dropped significantly even in permeable rocks containing permeable faults (Koziy et al., 2009).

Ore mineral characteristics

According to Cox et al. (2003), characteristic ore minerals formed in sediment-hosted Cu deposits are pyrite, chalcopyrite, bornite, different types of CuS-minerals as chalcocite and digenite, and minor amounts of sphalerite and galena. Silver in native form is common, and copper in native form is found where the sulfide content is low. Some of these minerals were found in different zones of the deposits, as pyrite was found close to reduced rocks while chalcocite was found close to an oxidized source rock. Based on different ore minerals occurring at different places in the deposit, it can be divided into a high-grade copper zone and a low-grade copper zone. The high-grade zone is characterized by the appearance of chalcocite and bornite with pyrite as a trace element. The low-grade zone is characterized by appearances of pyrite and chalcopyrite, with a negative correlation of pyrite and the copper grade. For reduced-facies, mineralization is characterized by a zonation from chalcocite to bornite to chalcopyrite to pyrite towards the underlaying source rock of copper. Cox et al.

(2003) also mentioned that in 10 deposits, which nearly all of them are “important” deposits, they contain different arsenic minerals such as tennantite. These arsenic minerals occur together with chalcopyrite and represent a later hydrothermal overprint in the high-grade zone, overprinting earlier epigenetic mineralization of bornite-chalcocite (Cox et al., 2003).

(39)

Examples of sediment-hosted deposits

Well-known sediment-hosted copper deposits are the Permian Kupferschiefer deposits in Central Europe and the Neoproterozoic Zambian Copperbelt deposits in Congo, Africa. The oldest known sediment-hosted copper deposits are the Udokan deposit in Russia, with an age of 2.2-1.8 Ga, according to Abramov (2008) within Hitzman et al. (2010). Deposition of the sediments occurred between ⁓2.2 to 2.06 Ga and mineralization at ⁓1.9 Ga related to a collision between two domains that also metamorphosed (mostly at greenschist facies) and inverted the metasedimentary basin according to Perello et al. (2016).

3.2 Volcanogenic massive sulfide deposits (VMS)

Figure 12: (A) Cross-section of a mid-oceanic ridge environment with formations of black smokers due to fluid circulation, with characteristics of fluids also listed. From Laurence (2020, p. 197) based originally on Lydon (1998) and Scott (1997); (B) Lenticular shaped VHMS deposit with ideally structure and associated ore

mineralization from Sawkins (1976). Cpy: Chalcopyrite, Py: Pyrite, Po: Pyrrhotite, Sp: Sphalerite, Gn: Galena, Ba:

Barite, Hem: Hematite.

VMS deposits, also called volcanic hosted massive sulfide (VHMS) deposits are deposits typically enriched in Cu and Zn that have been formed from hydrothermal fluids circulating in volcanic active submarine environments (Laurence, 2020, p. 195). Black and white smokers are undersea chimneys associated with VMS deposits forming today, especially around mid- oceanic ridges (Figure 12A). A magma source related to volcanic activity heats the fluids as they travel down towards the magma, elements in the surrounding rocks are picked up, and fluids start to ascend at a certain temperature and pressure. On the way up, the hot

hydrothermal fluids are concentrated in channels within the stockwork zone shown in Figure 12B above and escape through chimneys into seawater (Laurence, 2020, p. 196, 197).

Precipitation of minerals occurs mainly due to a drop in temperature on the way up, building the chimney itself, the mound around by settlement of particles and chimney collapse, and

(40)

mineralization occurs by replacement processes within the mound itself (Lydon, 1998). This mound is, according to Sawkins (1976), characterized by a stratiform zonation of minerals as different elements have a temperature-dependent solubility. This temperature does usually evolve from colder to hotter temperatures as the deposit evolves, with Cu being significantly more soluble at higher temperatures from around 250°C, and Fe in pyrite at even higher temperatures (Laurence, 2020, p. 201, 202). The mineral assemblages in Figure 12B above shows this zonation, with chalcopyrite, pyrite, and pyrrhotite forming the base of the deposit (at a later stage), above the stockwork zone which contains disseminated sulfides (Sawkins, 1976).

Ancient VMS deposits have been exposed to different plate tectonic processes, which most likely have destroyed chimneys, but associated mounds of a stratiform massive sulfide mineralization can be preserved (Figure 12B; Sawkins 1976).

3.3 Stable isotope

Isotopes are atoms with the same amounts of protons but different amounts of neutrons.

Stable isotopes are isotopes of an element that does not change the number of neutrons over time. In contrast, unstable or as also called radiogenic isotopes, do change this number of neutrons over time. As neutrons have a significant weight in atoms, different isotopes of the same element have different weights (Hoefs, 2018, p. 1).

The different weights of isotopes make the lighter isotopes more readily to interact in

chemical processes than heavier isotopes as less dissociation energy is needed to break bonds due to the vibrational frequency and energy are higher in lighter isotopes (Tiwari et al., 2015, p. 68). Different geological processes can affect the distribution of the heavier and lighter isotopes of the same element between coexisting phases by isotopic fractionation, and therefore can stable isotope ratios be used to trace back those processes. The two main types of isotopic fractionation processes are, according to Tiwari et al. (2015, p. 68), equilibrium fractionation and kinetic fractionation.

Equilibrium fractionation

The equilibrium fractionation implies that there is an equilibrium between the light and the heavy isotopes in coexisting phases, and changes in this equilibrium are only caused by isotope exchanges (Tiwari et al., 2015, p. 68, 69). The equilibrium depends mostly on the temperature, with higher temperature giving less fractionation and opposite, but it is also to

(41)

some degree depending on the pressure in the system (Hoefs, 2018, p. 6). Evaporation- condensation processes are an example that can lead to variation in isotopic composition by pressure change, as lighter isotopes will preferentially be in the vapor phase and heavier in the liquid phase (Hoefs, 2018, p. 8, 9). Formation of carbonates enriched in ¹³C is usually caused by equilibrium fractionation as inorganic carbon from the atmospheric CO2 enters the

seawater, as seen in the equations below (Hoefs, 2018, p. 62).

𝐶𝑂₂(𝑎𝑞. ) + 𝐻₂𝑂 ↔ 𝐻₂𝐶𝑂₃(𝑎𝑞. )

Equation 3: Carbon dioxide reacts with H2O forming carbonic acid in equilibrium (Hoefs, 2018, p. 62).

𝐻₂𝐶𝑂₃(𝑎𝑞. ) ↔ 𝐻⁺+ 𝐻𝐶𝑂₃⁻(𝑎𝑞. )

Equation 4: Carbonic acid forming H⁺ and bicarbonate in equilibrium (Hoefs, 2018, p. 62).

𝐻𝐶𝑂₃⁻(𝑎𝑞. ) ↔ 𝐻⁺+ 𝐶𝑂₃²⁻(𝑎𝑞. )

Equation 5: Bicarbonate forming H⁺ and carbonate ion in equilibrium (Hoefs, 2018, p. 62).

𝐶𝑎²⁺(𝑎𝑞. ) + 𝐶𝑂₃²⁻(𝑎𝑞. ) → 𝐶𝑎𝐶𝑂₃(𝑠)

Equation 6: Calcium and carbonate ions forming calcite (Hoefs, 2018, p. 62).

What also can be formed in addition to calcite is dolomite if magnesium (Mg²⁺) is available in the system or ankerite if an addition of iron (as Fe²⁺) and manganese (Mn²⁺) are available.

Kinetic fractionation

The kinetic fractionation can also cause isotopic fractionation and is mainly based on the rate of reactions in molecules that forms no isotopic equilibrium (Tiwari et al., 2015, p. 69).

Diffusion is one process that can give rise to kinetic fractionation as there will be a fractionation between light and heavy isotopes, as lighter isotopes will travel faster than heavier ones (Hoefs, 2018, p. 10, 12). Organic carbon is typically isotopic light (⁓ -25‰) enriched in ¹²C by kinetic fractionation effects during complex reactions when carbon dioxide is transformed into oxygen by photosynthesis in both terrestrial and aquatic plants (Hoefs, 2018, p. 64-66). Marine carbonates have in general a δ¹³C value of ⁓0‰, which can be disturbed by infiltration of light organic carbon if production and burial of organic matter are high. The organic matter enriches the marine carbonates in lighter ¹²C, which moves the δ¹³C into a lighter composition (Hoefs, 2018, p. 65, 66).

Geochemical characteristics of sediment- and mafic rock-hosted Cu deposits in the Kåfjord area, Alta-Kvænangen Tectonic Window, Northern Norway